AI at Work◆ AI Tip
The Model Gets Replaced. The Manual Doesn't.
I had the strongest model on my account write the operating manual its replacement runs on. Eight procedures. Any model. Yours to steal.
July 7, 20265m read
AI
Claude Fable Dropped. The Second Brain Is Still the Point.
Anthropic's most powerful model just shipped, and the most instructive reaction came from someone who didn't rebuild anything. Nate Herk spent the day slotting Fable into the AI OS he'd already built — folders, markdown, and the Four Cs. That's the lesson.
June 10, 20267m read
AI◆ Deep Dive
10 Finance + AI Repos, One Signal
This week's fastest-growing finance + AI repos on GitHub. Strip the names away and one pattern is left standing: the whole field is converging on multi-agent architectures and MCP.
May 31, 20266m read
Engineering
Building an Agent Operations Platform in One Session
One YouTube video, one insight, one session. What started as a note about persistent agent expertise turned into a full agent operations platform: 22 specialist roles, behavioral health monitoring, directive lifecycle tracking, and an E2E proof that closed in 8 minutes.
March 30, 202612m read
AI
53 Hours a Day: What AI Agent Orchestration Actually Looks Like
I produced 1,060 hours of verified engineering output in 20 days. Not by coding faster — by commanding AI agents in parallel. Here's the audit trail.
March 28, 20268m read
AI
Perplexity Computer and the Fork in the AI Agent Road
Perplexity just launched a managed multi-model agent platform that orchestrates 19 AI models. It is a direct shot at open-source agent systems — and the architectural trade-offs tell you everything about where this industry is splitting.
February 27, 20268m read